The link (Appendix) provides the appendix file for the ICDM 2018 paper: On Multi-Query Local Community Detection.
The link (Code) provides the code (implemented with C++) we used in the paper.
Here is a MATLAB version.
The datasets we used are from (Note: please refer to associated original papers):
1. Networks with ground-truth communities: Amazon, DBLP, YouTube, LiveJournal, Orkut
Co-purchase network: Amazon;
Collaboration network: DBLP
Social network: YouTube, LiveJournal, Orkut
2. Brain Co-activation Network
This brain co-activation network consists of 638 nodes which correspond to the cortical areas of the human brain and 18625 edges which measure the functional associations between the cortical areas.
The flavor compound network represents the compound similarity between different ingredients. Nodes in the flavor compound network represent ingredients. There is an edge between two ingredients if they share similar flavor compounds.
If it is useful for your projects, please kindly consider citing our papers.
@INPROCEEDINGS{MRW,
title={On Multi-query Local Community Detection},
author={Yuchen, Bian and Yaowei, Yan and Wei, Cheng and Wei, Wang and Dongsheng, Luo and Xiang, Zhang},
booktitle={IEEE International Conference on Data Mining (ICDM)},
year={2018},
pages={9-18}
}
@ARTICLE{MRW_KAIS,
title={Memory-based random walk for multi-query local community detection},
author={Yuchen, Bian and Dongsheng, Luo and Yaowei, Yan and Wei, Cheng and Wei, Wang and Xiang, Zhang},
journal={Knowledge and Information Systems}
year={2019},
pages={1-35},
}